Normal Liver, Spleen, and Kidney

Dimensions in Neonates, Infants,

and Children: Evaluationwith
Sonography

OznurL.Konu@1 OBJECTIVE. The objectiveof this studywas to determinethe normal rangeof dimen
Ay@egüI
Ozdemir sions for the liver, spleen, and kidney in healthy neonates, infants, and children.
AlaaddinAkkaya SUBJECTS AND METHODS. This prospectivestudyinvolved307 pediatricsubjects
(169 girls and 138 boys) with normal physicalor sonographicfindingswho were examined
GoncaErba@
becauseof problems unrelated to the measuredorgans. The subjects were 5 days to 16 years
Had çelik
old. All measured organs were sonographically normal. At least two dimensions were ob
SedatI@lk tamed for each liver, spleen, and kidney. Relationships of the dimensions of these organs with
American Journal of Roentgenology 1998.171:1693-1698.

sex, age, body weight, height, and body surface area were investigated. Suggested limits of
normal dimensions were defined.
RESULTS. Dimensionsof the measuredorganswere notstatisticallydifferentin boysand
girls. Longitudinaldimensionsof all threeorgansshowedthe bestcorrelationwith age,body
weight, height, and body surface area. Height showed the strongest correlation of all. This
correlationwasa polynomialcorrelation.
CONCLUSION. Determinationof pathologicchangesin sizeof the liver,spleen,andkid
ney necessitates knowing the normal range of dimensions for these organs in healthy neonates,
infants, and children. Presented data are applicable in daily routine sonography. Body height
should be considered the best criteria to correlate with longitudinal dimensions of these organs.

S onography provides a quick as growth curves (who were not in the third to 97th
sessment of visceral organ di percentiles)were not includedin the study.Another
mensions without any risk of major criterion for selection of children was having
no clinically or sonographically pathologic findings
radiation. The normal range of visceral organ
related to the studied organs. Most children were
sizes in adults and children determined with
completely healthy. although some were undergo
sonography has been reported elsewhere [ I—¿
ing follow-up for a disease unrelated to the mea
13]. However, available data are limited for sured organs, such as hip dysplasiaor undescended
the liver and the spleen in children, which testes. No child had a history of oncologic, hemato
causes difficulty in defining hepatomegaly logic, or traumatic conditions. Some children with
and splenomegaly sonographically. urinarytract infectionwereincludedin the study,
Our purposewas primarily to document butonly liverandspleendimensionswererecorded
the normal range of dimensions of the liver inthosepatients. Wealsodidnotrecorddimensions
in children. The relationship of each dimen of a number of spleens for which abdominal gas

sion with sex,age, body weight, height,and distentionpreventedreliablesize measurements.All

measuredorganshada normalpositionandshape
body surface area was determined. A similar
and normalecho texture.
evaluation was done for the spleen and the
We used high-resolution real-time sonographic
ReceivedMarch30,1998;acceptedafterrevision kidney at the same time.
June17,1998. scanners (SSA 270A; Toshiba, Tokyo, Japan; and
EUB-5l5 and EUB-5l5A; Hitachi, Tokyo, Japan)
1All authors: Department of Radiology, School of Medicine,
GaziUniversity,Besevier,06510,Ankara,Turkey.Address Subjects and Methods with 3.5-MHz convex transducers.Patientshad
correspondenceto 0. L Konu@. neither preparation nor sedation.
AJR1998;1711693—1698 We prospectively examined 307 pediatric sub Liver measurements were performed in all sub
jects ( 169 girls and 138 boys) with sonography. The jects(n = 307).In a subjectlying in thesupineposi
0361—803X/98/1716—1693 rangeof agewasfrom full-termnewborns(5 days) tion, longitudinaland anteroposteriordimensions
©
AmericanRoentgenRaySociety to 16 years. Patients who did not have normal wereobtainedin the midclavicularandmidsagittal

AJR:171,December1998 1693
 Konu@etal.

planes for the right and left lobes. “¿Midsagiual the next zero or five integer values above the 95th
plane―means that the plane passes through the percentile, considering also the mean, minimum, Jver I Imenslons as
xiphoid process. In both planes, the upper margin of and maximum values of normal [I 1. Revealed by Sonography In
the liver was defined as the uppermost edge under I Pediatric Subjects
the dome of the diaphragm, whereas the lower mar
gin was defined as the lowermost edge of the lobe Results RightLobe Left Lobe
BodyParameter
(Fig. 1). Long. AP Long. AP
Spleen measurements were performed in 169 No statistically significant differences were
girls and 130 boys (n = 299). Longitudinal and found betweenthe two sexesin any agegroup Height.85.77.81.50Age.82.71.78.47Weight.80.73.74.49Surface
transverse dimensions in the coronal plane were for any measured organ dimension (t test, p>
obtained with the subject in the supine or slightly .05). Therefore, all data were rearrangedwith
rightlateraldecubitus
position.Longitudinal
size out being separatedaccording to sex. area.83.75.78.45
measurement was performed between the most su Compared with the anteroposterior and Note—Long.
=longitudinal,
AP=anteroposterior.
peromedial and the most inferolateral points of the transverse dimensions, all longitudinal di
spleen. The transverse dimension was measured
mensions of each organ showed the highest
between the hilum and the most superolateral mar
correlation with the body parameters(i.e.,
gin ofthe spleen (Fig. 2).
age, height, weight, and body surface area) ;orrelatlon Coefficientsfor
Kidney measurements were performed in 155
(Tables 1—3). Among the body parameters, pleen Dimensions as
girlsand 124 boys(Fl= 279). Longitudinaland Revealedby SonographyIn
transverse dimensions of both kidneys were ob height was the one best correlated with the 9 PediatrIcSubjects
tamedinthecoronalplanepassing
throughthere longitudinal dimension of each organ. Body
nal hilum with subjects in the supine or slightly surface area, age, and weight followed ParameterLongitudinalTransverseHeight
Body
rightor leftlateraldecubitus
positions(Fig.3). height, in that order. Correlation coefficients
American Journal of Roentgenology 1998.171:1693-1698.

To determine reproducibility. each measure are presented in Tables 1—3. On regression Age .84 .67
ment of each organ was repeated at least three analysis, the relationship between the longi Weight .84 .69
times, and the most repeated value was recorded.
tudinal dimensions and height was of the Surfacearea.88 .86.70 .70
Data obtained from the measurements were classi
polynomial type for each organ (Figs. 4—8).
fled into I I age groups. Relationships of all di
mensions with sex, age. height. weight. and body Longitudinal dimensions of the right lobe
surface area were statistically analyzed. Body sur of the liver, the spleen, and both kidneys,
face area was derived using the patient's height with corresponding values of patient height
__________Correladon Coefficients for I
and weight from the nomogram modified from the in selected age groups, are presented in Ta ,1q@@Kidney Dimensions as
dataof E. Boyd by C. D. WestI 141.All measure bles 4-7. Mean, minimum, and maximum Revealedby Sonographyin F
ments were plotted as a function of age in months, values; fifth and 95th percentile values; stan 279 PediatrIc Subjects j
body weight, height, and body surface area. Types dard deviations; and suggestedlimits of nor
of relationships and regression lines were obtained
@Long.
mal are also given. Other dimensions are BodyParameterRightKidneyLeftKidney
for the various sets of data using the least squares Trans.Long.Trans.Height.94
presented in graphic form (Figs. 4—8).The
method. The mean. minimum, and maximum
dimensions of the left lobe of the liver, ex .86.93.85E@Age.91
value; fifth and 95th percentiles: and standard de
viationswere calculated(Excel 5.0: Microsoft. cept those taken in infancy, poorly correlated .84.90.83Weight.89
Redmond, WA). The lowermost and uppermost with body parameters.
Significant differences were found between .82.88.79Surface
limitsof normalfor eachage groupwere sug
gested and were derived using the previous zero or the longitudinal sizes of the right and left kid .83.91.80Note—Long.
area.91
five integer values below the fifth percentile, and neys (t test, p < .05). For this reason, findings =longitudinal,
Trans.= transverse.

(7IuI:@ Posterior Anterior

@ Right Left

A B
Fig.1.—Diagrams
showhowlongitudinal
andanteroposteriordimensionsofliverlobesweremeasured. Fig.2.—Diagram
showshowIon Fig.1—Diagramshows
A,Livermeasurements
wereperformedatmidclavicularandmidlinesagittal
planes. gitudinalandtransverse
dimen how longitudinaland
B,Longitudinal
andanteroposterior
dimensions
of rightandleftlobesweremeasured. sionsof spleenweremeasuredin transverse dimensions
coronal section passing through of kidney were mea
splenichilum. suredincoronalsection
passingthrough renal
hilum.

1694 AJR:171, December 1998

 Sonography of Liver, Spleen, and Kidney Dimensions in Children

Fig.4.—Scatter
diagram organs without the risk of ionizing radiation.
shows longitudinal(•) Organ volumes or ratios obtained by using
and anteroposterior (0)
dimensions of right lobe various organ dimensions and body surface ar
of liver plotted against eas are already used in correlation with body
heights of patients. Note parameters to describe the normal dimensions
that longitudinal dimen
sions of right lobe are
andto measurethe degreeof pathologic devia
longer than anteroposte tion from normal. However, those techniques
nor dimensions and this are time-consuming and impractical in daily
difference increaseswith use. On the other hand, either one or more of
height.
the longitudinal, anteroposterior, and trans
versedimensionsof theseorgansaremeasured
aspart of routine abdominal sonographicscan
120 200 ning. Therefore, use of these measurements
Height(cm) seems more practical for purposes of deter
mining “¿normal:'
Any other data, when neces
sary to combine with the above measurements,
for the right and left kidneys are presentedsep include enough children to separate the sub should also be easily obtainable. Age, body
arately in Tables 6 and 7 and in Figures 7 and 8. jects into appropriate age groups [1—9].We weight, and height are suchparameters.
studied many children of all ages, and to our We did not find any significant difference
Discussion knowledge, this study covers the largest se in measured organ sizes between the two
In the literature, studies about normal liver ries of pediatric liver and spleendimensions. sexesof any age group (t test, p > .05). This
American Journal of Roentgenology 1998.171:1693-1698.

and spleen dimensions in children are rare, Sonographyis a simpleandreliableway to finding is similar to the findings of others [2—
especially for the liver. Most studies do not visualize and to measure abdominal visceral 7, 10—12]. Therefore, sex certainly is not a

j—-••—@——.@—100
120

@
@ .-@. . . 4 ..—

0 20 40 60 80 100 120 140 160 180 200

(cm)Hei
ght Height(cm)

Fig.5.—Scatter
diagramshowslongitudinal(•)
andanteroposterior
(o) dimen Fig.6.—Scatter
diagram showslongitudinal (•)
andtransverse
(0) dimensions
of
sions of left lobe of liver plotted against heights of patients. Note that longitudinal spleenplottedagainstheightsof patients.
dimensions of left lobe are shorter than anteroposterior dimensions at patient
heights of less than 70 cm but longer at patient heights of morethan 70 cm.

E
E
a..
a,
C
•¿C

.C
@,

0 20 40 60 80 100120Height 140 160 180 200 0 20 40 60 80 100 120 140 160 180 200
(cm) Height(cm)

Fig.1.—Scatter
diagram
showslongitudinal
(•)
andtransverse
(o) dimensions
of Fig.8.—Scatter
diagramshowslongitudinal(•)
andtransverse(0) dimensionsof
right kidney plotted against heights of patients. left kidneyplottedagainstheightsof patients.Leftkidneydimensionsarelonger
thanrightkidneydimensions, asseenin comparison withFigure7.

AJR:171,December1998 1695
 Konu@etal.

ILongitudinal
Dimensions
ofRight
Lobe
ofUver
Versus
Height
and
Age
Subjects LongitudinalDimensions(mm)of RightLobeof Liver
Normal5th95thLowermostUppermost47—64
BodyHeight Range . Limitsof
Minimum. MaximumPercentileSuggested
(cm)No.Age (mo)MeanSD.

54—73 40 4—6 73 10.8 44 92 53 86 45 95

65—78 20 7—9 79 8.0 68 100 70 90 60 100
71—92 18 12—30 85 10.0 67 104 68 98 65 105
85—109 27 36—59 86 11.8 69 109 63 105 65 115
100—130 30 60—83 100 13.6 73 125 77 124 70 125
110—131 38 84—107 105 10.6 81 128 90 123 75 130
124—149 30 108—131 105 12.5 76 135 83 128 75 135
137—153 16 132—155 115 14.0 93 137 95 136 85 140
143—168 23 156—179 118 14.6 87 137 94 136 85 140
152—17553 121—3 180—20064 12110.4
145L.ongitu4l 11.745 10090 14148 10482 13940 9590
American Journal of Roentgenology 1998.171:1693-1698.

Dimen‘ons ofSplei Versus Height and Age

SpleenBody SubjectsInal LongitudinalDimensions(mm)of

Normal5th95thLowermostUppermost48—64
Height Range . Limitsof
Minimum. MaximumPercentileSuggested
(cm)No.Age (mo)MeanSD.

54—73 39 4—6 59 6.3 45 71 47 67 40 75

65—78 18 7—9 63 7.6 50 77 53 74 45 80
71—92 18 12—30 70 9.6 54 86 55 82 50 85
85—109 27 36—59 75 8.4 60 91 61 88 55 95
100—130 30 60—83 84 9.0 61 100 70 100 60 105
110—131 36 84—107 85 10.5 65 102 69 100 65 105
125—149 29 108—131 86 10.7 64 114 70 100 65 110
137—153 17 132—155 97 9.7 72 100 81 108 75 115
143—168 21 156—179 101 11.7 84 120 85 118 80 120
152—17552 121—3 180—20053 1017.8 10.333 8871 12040 8865 11530 8570 120

@I•LongltudDimen@ronsof RightKidney VersusHeight

KidneyBody Subjects@nal
and
AgeLongitudinal
Dimensions(mm)
of Right
Normal5th95thLowermostUppermost48—64501—3505.838664058356554—73394—6535.341665064407065—78177—9595.250705266457071—921812—30613.4
Height Range .
Limits of
Minimum. MaximumPercentileSuggested
(cm)No.Age (mo)MeanSD.

1696 AJR:171,December 1998

 Sonography of Liver, Spleen, and Kidney Dimensions in Children

•¿I@1:1@@LongItudInaI
AgeSubjectsLongitudinal
Dimensionsof Left KidneyVersus Height and
KidneyBody Dimensions (mm) of Left

Normal5th95thLowermostUppermost48—64
Height Range -
Limitsof
Minimum. MaximumPercentileSuggested
(cm)No.Age (mo)MeanSD-

54—73 39 4—6 56 5.5 44 68 47 64 40 70

65—78 17 7—9 61 4.6 54 68 54 68 45 75
71—92 18 12—30 66 5.3 54 75 57 72 50 80
85—109 22 36—59 71 4.5 61 77 61 76 55 85
100—130 26 60—83 79 5.9 66 90 70 87 60 95
110—131 32 84—107 84 6.6 71 95 73 93 65 100
124—149 27 108—131 84 7.4 71 99 75 97 65 105
137—153 15 132—155 91 8.4 71 104 77 102 70 110
143—168 22 156—179 96 8.9 83 113 84 110 75 115
152—17550 111—3 180—20050 995.5 7.539 8761 11642 9059 11035 8065 120

determining factor for organ dimensions in direct roentgenography in 350 children to es In the study of Markisz et al. [4] that in
American Journal of Roentgenology 1998.171:1693-1698.

the pediatric age group.
In most other studies, sizes between the
fifth and the 95th percentile were the ac
cepted normal limits [2, 3, 5, 6, 8—10,13].
However, this practice results in approxi
mately 10% of children with normal visceral
organ sizes falling outside these limits [1 1].
Besides, although plus or minus two SDs are
the accepted equivalents of the fifth and 95th
percentile values statistically [15], some
studies were based on plus or minus one SD
[7]. For this reason,we preferredto define
the normal lowermost and uppermost longi
tudinal dimensions of the studied organs us
ing the method originally described
Rosenberget al. [ 11in 1991.
We foundthatheightshowsthebestcorre
lation with any one of the mentioned organ
dimensions. Body surface area also shows a
high correlation with organ dimensions, but
to a lesser degree. However, its derivation is
not as practical as height. For those reasons,
we preferred not to use body surface area in
standard tables and graphics. Age and body
weight are not as important as height and
body surface area.Therefore, when deciding
if sonographically obtained dimensions of an
organ are normal, patient height should be
the primary concern.
The liver is an organ of complex shape
and varies widely from patient to patient.
Our wide range of normal dimensionssup
ports this opinion. However, it is necessary
to have references for normal liver dimen
sions in the pediatric age group.
The findings of various similar studies dif
fer somewhat from each other and from ours.
Deligeorgis et al. [12] performed a study with
tablish normal liver dimensions. They found cluded 116 children, liver and spleen dimen
significantly longer dimensions than we did. sions obtained with scintigraphic methods
The reason for the difference is obviously the also exceeded ours. Those authors used vol
magnification factor on radiographs. ume measurements and found a high linear
Somefindings are contradictoryeven within correlation between the volumes of these or
onestudy.Forexample,in the studyof Holder et gans and patient weights. Correlations with
al. [2], which included 185children,sonograph age were lessevident. Patient height or organ
ically measureddimensionsof the liver are an dimensions were not consideredin their study
averageof 4 cm smaller than the samedimen and limits of normal cannot be derived. From
sions measured scintigraphically. Our sono the study of Markisz et al. (like the study of
graphicallymeasuredliver dimensionsarelarger Holder et al. [2]), we conclude that scinti
than those of Holder et al. The reason is proba graphically obtained organ sizes are larger
bly their techniqueof measurement.They used than sonographically obtained ones. There
linear transducersand performed the measure fore, differences in organ size are probably
by ments in a sagittal plane passing through the due to the difference of imaging techniques.
midpoint of the liver's right and left margins. In 1983, Niederau et al. [13] described a
Linear transducers,becauseof the interposition study of normal liver dimensions in 1000
of lung betweenthe dome of the liver and the adult patients. Sonographically, their study
anteriorabdominalwall, preventsomeof the su showed smaller values than ours. This differ
perior portion of the liver from being observed. ence is probably the result of the same tech
This effect,asHolder admits,resultsin the “¿obnical reason we discussed with respect to the
served― lung—liver
border'sbeing determinedas study of Holder et al. [2]. Niederau et al.
the superior margin of the liver, and conse noted that organ sizes increased with height
quently in the liver's being measuredas having and body surface area in adults.
smaller longitudinal dimensions than the livers In 1983, Dittrich et al. [3] also used
we measured.Convex probes,such as the one sonography to study liver dimensions in 194
we used, prevent this complication. Besides, un children. Although they reported smaller di
like Holderet al.,we believethatthemidpoint of mensions than we did, Dittrich et al. noted
the liver's right and left margins cannot be re the best correlation was found between liver
ferred to as a standard sagittal plane because the dimensions and height, which is a linear cor
left lobe in particular differs in extensionand relation. They found also a high correlation
size from one personto anotherand with age. with body surface area, but in their study
Instead, midsagittal and midclavicular planes only height was used as a reference parame
seem more appropriate for longitudinal liver size ter for further analysis.
measurements. In spite of these technical differ Normal spleen dimensions and volume
ences, Holder et al. also concluded that height is standards have been investigated by only a
the best-correlatedbody parameterwith liver few researchers.Rosenberg et al. [1] studied
longitudinal dimensions. normal spleen longitudinal dimensions in

AJR:171,December1998 1697
 Konu@et al.

230 children using sonography.Interestingly, 8). This finding is similar to the findings of should be compared with tables and graphics
their dimensions are smaller than ours for Dinkel et al. [10]. In the infancy period, the of normal parameters.
children under the age of 8 years, whereas in anteroposterior dimension of the left lobe of
children over the age of 8 years, the findings the liver may be somewhat longer than the
of their study and ours are similar. They longitudinal dimension. With aging, these References
stated that body weight showed the best cor phenomena change and the longitudinal di 1. Rosenberg HK, MarkowitzRI, KolbergH, Park
relation with spleen length, which was mension of the left lobe becomes longer than C, HubbardA, BellahRD. Normalsplenicsizein
roughly logarithmic in type. Height and age, the anteroposterior dimension except in infantsand children:sonographicmeasurements.
AiR 1991;157:119—121
in that order, were less well correlated with obese persons [13]. We observed the same
2. Holder L, Strife J, Padikal TN, Perkins PJ, Ker
spleen length. phenomenon in infants. Anteroposterior di
elakesJG. Liversize determinationin pediatrics
mensionsof the left liver lobeswere foundto
Schlesinger et al. [9] studied normal spleen using sonographic and scintigraphic techniques.
volume in 48 children using CT. These re be somewhat longer than the longitudinal di Radiology1975;!17:349—353
searchersfoundthatspleenvolumecorrelated mensions. The reason is that the intraabdom 3. Dittrich M, Milde S. Dinkel E, Baumann W,
better with body weight than with age. The inal visceral organs are larger in children Weitzel D. Sonographic biometry of liver and
best regressionmodel was a simple linear re than in adults when compared with total spleen size in childhood. Pediatr Radiol 1983;13:
lationshipof spleenvolumeto bodyweight. body volume. This finding leads to the con 206—211
4. Markisz JA, TrevesST.Davis RT Normal hepatic
Renal dimensions, as well as cross-sec clusion that measurements of the midclavic
and splenic size in children: scintigraphic deter
tional areas and volumetric standards of the ular and midsagittal longitudinal dimensions mination.PediatrRadiol 1987;17:273—276
kidneys, have been investigated by many dif suffice to estimate the liver size in most chil 5. Christophe C, Cantraine F, Bogaert C, et al. Ultra
ferent authors for pediatric age groups. dren, and that measurements of anteroposte sound: a method for kidney size monitoringin
Christophe et al. [5] measured kidney rior dimensions, particularly of the left lobe, children. EurJ Pediatr 1986;145:532—538
American Journal of Roentgenology 1998.171:1693-1698.

sizes sonographically in 170 children. Values must be obtained only in infants. 6. Han BK, Babcock DS. Sonographic measure
of longitudinal kidney dimensionsin their Data of normal visceral organ sizes accord ments and appearance of normal kidneys in chil
dren.AJR1985;145:611—616
study are also similar to ours. These re ing to age have been given in table form in
7. Dremsek PA, Kritscher H, BöhmG, Hochberger
searchers found a linear relationship between several previous studies [1, 7]. However, the
0. Kidneydimensionsin ultrasoundcomparedto
renal length and height. Additionally, they selected age groups in those studies present a somatometric parameters in normal children. Pe
found that the longitudinal dimension of the wide range like 5 years, as in the studies of diatrRadiol 1987;17:285—290
kidneys versus body surface area yielded the Rosenberg et al. [1] and Dremsek et al. [7] 8. Rosenbaum
DM, KorngoldE, TeeleRL. Sono
most accurate correlation. However, as they (whereas our selected age groups had an age graphic assessment of renal length in normal chil
stated, correlating renal length with body range of 3 months or 2 years). In other studies, then.AJR1984;142:467—469
height is more practical. data were given only in graphic form [2—6, 9. SchlesingerAE,EdgarKA,BoxerLA.Volumeof
the spleenin childrenas measuredon CT scans:
In 1985, Dinkel et al. [10] studied renal di 10]. We present our data both in tabular and
normal standards as a function of body weight.
mensions and volumes in 325 children with graphic forms with the aim ofenabling a more AJR 1993;160:1107—l 109
sonographyandprovidedseparategraphicsfor practical evaluation during a sonographic ex 10. DinkelE, ErtelM, DittrichM, PetersH, BerresM,
the right and left kidneys. Correlations of renal amination. Schulte-Wissermann H. Kidneysizein childhood:
dimensionswith body parametersin that study Establishing normal parametersis manda sonographic growthchartsfor kidneylengthand
are similarto our findings,and thoseauthors tory for defining the pathologic changes in volume. PediatrRadiol 1985;15:38—43
11. Zerin JM, Blanc CE. Sonographicassessmentof
statedthat renal length showeda high correla size of the liver, spleen, and kidneys in rou
renal length in children: a reappraisal. PediatrRa
tion with body height. However, in their study, tine sonographic examinations of children.
diol 1994;24:l0l—106
the correlation was a linear relationship. These The methodsof measurementand analysis 12. Deligeorgis D, YannakosD, Doxiadis S. Normal
researchers also found that renal volume we used in this study are standardizedand size of liver in infancy and childhood: x-ray
showed the best correlation with body surface easy to apply. Findings are handy and reli study.ArchDis Child1973;48:790—793
area and showedalmostthe samecorrelation able and are suitable particularly for sonog 13. Niederau C, Sonnenberg A, Muller JE, Ercken
with body weight. For practical purposes,how raphy units and pediatric departments with brecht JF, ScholtenT, FritschWE Sonographic
large numbers of patients. measurements of the normal liver, spleen, pancreas,
ever, they proposed to use the correlation of re
and portal vein. Radiology 1983;149:537—540
nalvolumewith bodyweight. Longitudinal dimension and patient height
14. Berhman RE, Kliegman R. Nelson essentialsof
We observed that the increase in the longi are the most important determining factors of pediatrics. Philadelphia: Saunders, 1990:712
tudinal dimensions of these organs is much organ size among studied dimensions and 15. Remington 1W, Schork MA. Statistics with appli
more rapid during the first years of life, body parameters. Therefore, any specific cation to the biological and health sciences, 2nd
which mirrors the accentuated body growth longitudinal dimension should primarily be ed. Englewood Cliffs, NJ: Prentice Hall, 1985:
during infancyand early childhood(Figs. 4— correlated with patient height, and findings 23—43

1698 AJR:171,December1998